Sony Music and UMG Sue Suno Over Alleged Copyright Infringement in AI Models
Labels claim Suno's v6 AI models still use unlicensed music despite new partnerships.
1Core copyright allegations
Sony Music and UMG allege Suno's v6 AI models were trained on unauthorized music, despite the company's licensing deals with Warner and BMG. The labels specifically challenge Suno's admission that v6 incorporated 'user interactions' with prior models, which they interpret as outputs from illegally trained earlier versions. This forms the basis for their $9 billion damages claim under US copyright law, plus additional penalties for alleged YouTube scraping.
2The $9 billion question
The lawsuit calculates potential damages based on 60,202 allegedly infringed sound recordings. US copyright law sets statutory damages at up to $150,000 per willful infringement, leading to the massive total. Additionally, the labels seek $2,500 per instance for circumventing YouTube's download protections. These figures represent an existential threat to Suno, dwarfing typical AI copyright settlements.
3Suno's defense strategy
Suno maintains its v6 models were trained properly using licensed content from partners WMG, BMG, and Believe. The company argues user interactions and preference signals constitute fair use, not copyright violations. Suno's statement emphasizes its mission to democratize music creation through AI, positioning itself as an industry collaborator rather than infringer.
4Prior legal history
This lawsuit follows a 2026 data breach revealing Suno scraped millions of songs from YouTube Music, Deezer and Genius for earlier models. The company previously defended such scraping under fair use doctrine. Notably, Suno discontinued support for these older models when launching v6, suggesting strategic distancing from past practices that drew industry criticism.
5Industry implications
The case tests boundaries of AI training data legality in creative industries. Record labels increasingly view unlicensed AI training as commercial threat, while AI firms argue for transformative fair use. Suno's licensing deals with some majors but not others creates a fragmented legal landscape where partnerships don't guarantee comprehensive protection from lawsuits.
6Technical training details
Suno's v6 models combine three data sources: licensed partner content, user-generated creations/preferences, and internal team learnings. The legal dispute centers on whether user interactions indirectly incorporate unlicensed material from previous models. This technical nuance could determine whether courts view v6 as sufficiently distinct from allegedly infringing predecessors.
7Strategic timing
The lawsuit arrives as Suno attempts to position v6 as an industry-compliant product following criticism of earlier models. Labels appear determined to challenge even AI systems with partial licensing, signaling unwillingness to accept any unauthorized training regardless of subsequent partnerships. This suggests ongoing tension between AI innovation and copyright enforcement.
8Broader AI music landscape
The case exemplifies growing conflicts between generative AI and content industries. While Suno seeks to frame AI as strengthening the music ecosystem, labels view unchecked training as undermining copyright value. The outcome could establish precedents affecting all AI music tools, potentially requiring complete licensing or technological solutions to prevent unauthorized data incorporation.